Vast applications of silicon nanoclusters in various fields of industry, especially modern advanced technologies, have put a spotlight on these nanoparticles for the last three decades. In this thesis we tend to find stable structures of silicon nanoclusters, Si n (n=8-80), to be able to investigate their structural and electronic properties. This purpose is served by using the evolutionary algorithm implemented in the USPEX code which is then interfaced with the semi-empirical DFTB+ code. Afterwards, we further optimize some of the more stable structures by the means of the all-electron full-potential FHI-aims code at the DFT level to verify the results. Our results show that there is a geometrical transition from small- to medium-sized clusters at n=27. Plus, we introduce some new leading candidates for global minimum in the size range of n=20-60 atoms. Our method is computationally very efficient which acts as a filtering mechanism to quickly lead us to possible candidates for global minimum.
